Topics in GC & HPLC

Acheiving the sensitivity requirements for the Class 1 solvents in the revised USP<467> can be, let's face it, just plain difficult. By far, the biggest troublemaker is carbon tetrachloride in Procedure A. Sometimes a signal to noise ratio (S/N) of merely 3 seems a bit out of reach. Lets consider the concept of S/N for a moment. A S/N is just as it is desicribed - a measure of the analyte signal in relation to the noise of the system. Our first attempt, especially in headspace, is almost always to increase signal - more mass on column. But in the case of USP<467>, a defined method, we should first lower our noise. USP<467> is essentially an isothermal FID analysis, where peak widths are above 0.1 minutes. A data acquisition rate of 5hz can greatly reduce the noise of the system, and still maintain data quality. When comparing a 5hz to a 50hz rate, we can see the reduction of noise, not the increase in signal, was the difference between pass and fail.

FID is primary a detector that “sees” hydrocarbon type components. FIDs are insensitive to H2O, CO2, CS2, SO2, CO, NOx, and noble gases because they are not able to be oxidized/ionized by the flame. However: Sometimes a small signal is observed when a big air peak passes, caused by a small change of sensitivity (the ratio air : H2 changes when the peak elutes).

Bigger impacts have been seen if columns become active. Air can indirectly be detected as it can react with stationary phase forming a bleed product that is detected by FID.

If systems are shut-down without positive flow, while the detector is still hot, back diffusion of air from the detector will happen, oxidizing and hydrolizing the last 10-cm of column, which is in the hot zone. This “activated” part can cause big trouble by adsorption of analytes and also by acting as a catalyst for phase degradation, see example.

Key learning: Always keep positive flow until the detector/injector had time to cool.

Direct injection is obtained when all the sample injected in the liner is evaporated instantly and directly sent to the column. If the column is set at a low temperature, focusing can be obtained for components with higher retention. The solvent peak will usually be very broad. To eliminate condensation effects for the solvent, the column must be set at a temperature where the solvent will be in gas phase, so usually near or above the boiling point of the solvent.

Direct injection is done : (1) if we need to measure sub ppm amounts in a matrix which is problematic with Splitless injection (2) if Splitted injection is discriminating to much. Special Uniliners are available which make a perfect seal for direct injection.

for best possible result: - Inject fast and smallest possible amount - Use 0.32/0.53mm columns for high flow - Use 2 mm or < liner ID - use column with thick film for retention Applications: traces of volatiles in high boiling and/or polar solvent

I was at the Recent Advances in Food Analysis conference in Prague last week and presented work on analyzing pesticides in dietary supplements, a very important subject given that FDA is moving to regulate these products. They are very complex herbal, botanical, etc. materials, which makes looking for trace level components very tricky. My colleagues, Jason, Julie, and Michelle tackled the sample preparation with QuEChERS and a cartridge SPE (GCB/PSA) cleanup. Even then, the resulting extracts had lots of peaks, many coeluting.

We then applied GCxGC-TOFMS with Rxi-5Sil MS and Rtx-200 columns to determine pesticides. Recoveries of over 40 pesticides, including organo- chlorine, nitrogen, and phosphorus, were very good and the GCxGC approach allowed unbiased quantification where single dimensional GC did not for numerous compounds.

Check out the contour plots (GCxGC chromatograms) of clean (!) extracts.

I’ve used ezGC Modeling Software since 1997 and have heard quite a bit of dialog on its utility. ezGC allows a user to enter two different retention times on a given column and use that data to adjust: carrier gas type, flow (constant flow / constant pressure), temperature, column length, film thickness, column ID, and pre-column (guard column)...this can be done simultaneously. It also has advanced features that allow adjustments to peak widths & response that can replicate what you would expect to see on your GC. The actual GC runs below give you an idea of the utility of ezGC; a slow program would not work, a fast program has coelutions and the modeled ezGC suggested solution works. I have attached a link to the whole chromatogram and it’s clear it would take a lifetime to get these analytes resolved without this program. Yes there are a few bugs here and there, but once you realized the absolute power of this program you will see it as Heaven-sent. (Cat#21487)

Of Restek’s seven variations of PEG phases the top sellers are the Stabilwax & Rtx-Wax columns. While the Stabilwax phase has the advantage of low bleed and stability; the ability of the Rtx-Wax to operate at below 40°C is especially advantageous for the analysis of low molecular weight volatiles. This is true for concentration techniques involving HS or P&T analytes which must be cold trapped onto the head of the column from the gas phase. Many PEG phases (including Stabilwax) undergo a solid-liquid phase transition T(g) below 40°C that results in loss of capacity and poor retention time reproducibility. Unfortunately the example shown here is my personal mis-use of the Stabilwax Phase–so unless you want to add to my blog choose the right PEG for your specific analysis and don’t hesitate to ask us for help.

Hi folks:

It has been a while since my last post, as our Web Master Kent Rauch has kindly (?) pointed out, but I have a good excuse. I've been in the lab exploring a variety of GC injection techniques. And as I work, I frequently rely on Grob's book "Split and Splitless Injection for Quantitative Gas Chromatography", an excellent read. If you don't have time for a book, get some quick injection pointers from Grob, as Restek recently reposted essays from "Koni's Korner". Some select (and timely!) titles include:

Why Uncoated Capillary Precolumns Enable Injection of Large Volumes

Sample Vaporization in Hot GC Injectors

Certification of injectors and injection techniques? Comments on splitless injection by readers.

Check these out and let me know if you have any comments, or want to discuss your own injection techniques.

P.S. Those that attended my splitless injection talk at the Florida Pesticide Residue Workshop know exactly what the Headline means.

Method 524.3 was approved by the EPA and promulgated; meaning that it has been published in the Federal Register & can replace 524.2 Revision IV. Poor purging compounds were removed from the target list (i.e. 2-nitropropane, nitrobenzene, 1,1-dichloropropanone); Chlorodifluoromethane (Freon® 22) was added as an emerging contaminant; and Oxygenates were added. HCl preservation was removed and P&T modifications are allowed; including heated purge. Single Ion Monitoring (SIM) has also been added. The method was developed using an Rtx-VMS 30m x 0.25mm x 1.4df Cat#19915 with Tekmar equipment, however, the method allows other columns & instrumentation to be used. OI Analytical has demonstrated that the Eclipse 4660 P&T using an Rtx-624 30m x 0.25mm x 1.4df cat#10968 is also appropriate for the method (OI Publication 32620409). If you have any questions or need help with your drinking water method send me an e-mail.

method http://www.epa.gov/safewater/methods/pdfs/methods/met524-3.pdf

“Make up” gas is a gas flow that is used to sweep components through a detector to minimize band broadening. For FID often N2 is used, which is set at a flow of 10-20mL/min. Depending on the GC-brand and detector design, the use of make-up gas can improve sensitivity: check the manuf. recommendations. If you want to optimize make up gas flow for maximum sensitivity, inject a fixed amount of component, using different make-up gas flows. Measure the signal to noise of a component (do NOT look at the peak response only !).

Make up gas needs to be filtered using charcoal filters , same as the air and H2. Note that the signal of mass flow responding detectors, like FID, will not be affected by make-up gas, while the concentration – responding detectors, like a TCD, the signal is related to the dilution of the gas. If column flow is 2 mL/min, and make up gas is set at 8mL/min, we loose a factor 5 in sensitivity. That’s why with TCD-type detectors we need to minimize make up gas.

Sometimes we get customer reaction that porous polymer PLOT columns bleed a lot. They are used to the low bleed values using Rxi-series of columns and wonder why this is not the case with porous polymers. Porous polymer columns have a huge amount of stationary phase to generate sufficient loadability and retention. Typically porous layers are upto 20 um in thickness, That means for a 30m column, a porous polymer PLOT has approx. 200 x more stationary phase compared with a 30/0.25/0.25 type Rxi column.

Therefore much higher bleed values can be expected for the porous polymer type columns. If we aim for same bleed as Rxi, we must realize, the upper temperature should be listed 100C lower.

The max temp. listed for porous polymer columns is kept high, to be able to bake-out impurities that may accumulate on the column. The new generation Rt-Q-BOND columns do offer a lower bleed while films are thicker, reducing background but with higher capacity so technology is improving.